A Comparison of Methods for Studying the Tumor Microenvironment's Spatial Heterogeneity in Digital Pathology Specimens

The influence of resolution on the predictive power of spatial heterogeneity measures as biomarkers of liver fibrosis

Spatial heterogeneity of cells in liver biopsies can be used as biomarker for disease severity of patients. This heterogeneity can be quantified by non-parametric statistics of point pattern data, which make use of an aggregation of the point locations. The method and scale of aggregation are usually chosen ad hoc, despite values of the aforementioned statistics being heavily dependent on them. Moreover, in the context of measuring heterogeneity, increasing spatial resolution will not endlessly provide more accuracy. The question then becomes how changes in resolution influence heterogeneity indicators, and subsequently how they influence their predictive abilities. In this paper, cell level data of liver biopsy tissue taken from chronic Hepatitis B patients is used to analyze this issue. Firstly, Morisita-Horn indices, Shannon indices and Getis-Ord statistics were evaluated as heterogeneity indicators of different types of cells, using multiple resolutions. Secondly, the effect of resolution on the predictive performance of the indices in an ordinal regression model was investigated, as well as their importance in the model. A simulation study was subsequently performed to validate the aforementioned methods. In general, for specific heterogeneity indicators, a downward trend in predictive performance could be observed. While for local measures of heterogeneity a smaller grid-size is outperforming, global measures have a better performance with medium-sized grids. In addition, the use of both local and global measures of heterogeneity is recommended to improve the predictive performance.

Spatial modelling of the tumor microenvironment from multiplex immunofluorescence images: methods and applications

Spatial modelling methods have gained prominence with developments in high throughput imaging platforms. Multiplex immunofluorescence (mIF) provides the scope to examine interactions between tumor and immune compartment at single cell resolution using a panel of antibodies that can be chosen based on the cancer type or the clinical interest of the study. The markers can be used to identify the phenotypes and to examine cellular interactions at global and local scales. Several translational studies rely on key understanding of the tumor microenvironment (TME) to identify drivers of immune response in immunotherapy based clinical trials. To improve the success of ongoing trials, a number of retrospective approaches can be adopted to understand differences in response, recurrence and progression by examining the patient's TME from tissue samples obtained at baseline and at various time points along the treatment. The multiplex immunofluorescence (mIF) technique provides insight on patient specific cell populations and their relative spatial distribution as qualitative measures of a favorable treatment outcome. Spatial analysis of these images provides an understanding of the intratumoral heterogeneity and clustering among cell populations in the TME. A number of mathematical models, which establish clustering as a measure of deviation from complete spatial randomness, can be applied to the mIF images represented as spatial point patterns. These mathematical models, developed for landscape ecology and geographic information studies, can be applied to the TME after careful consideration of the tumor type (cold vs. hot) and the tumor immune landscape. The spatial modelling of mIF images can show observable engagement of T cells expressing immune checkpoint molecules and this can then be correlated with single-cell RNA sequencing data.

Putting infection on the map: Using heatmaps to characterise within- and between-host distributions of trematode metacercariae

The location of parasites within individual hosts is often treated as a static trait, yet many parasite species can occur in multiple locations or organs within their hosts. Here, we apply distributional heat maps to study the within- and between-host infection patterns for four trematodes (Alaria marcianae, Cephalogonimus americanus, Echinostoma spp. and Ribeiroia ondatrae) within the amphibian hosts Pseudacris regilla and two species of Taricha. We developed heatmaps from 71 individual hosts from six locations in California, which illustrate stark differences among parasites both in their primary locations within amphibian hosts as well as their degree of location specificity. While metacercariae (i.e., cysts) of two parasites (C. americanus and A. marcianae) were relative generalists in habitat selection and often occurred throughout the host, two others (R. ondatrae and Echinostoma spp.) were highly localised to a specific organ or organ system. Comparing parasite distributions among these parasite taxa highlighted locations of overlap showing potential areas of interactions, such as the mandibular inner dermis region, chest and throat inner dermis and the tail reabsorption outer epidermis. Additionally, the within-host distribution of R. ondatrae differed between host species, with metacercariae aggregating in the anterior dermis areas of newts, compared with the posterior dermis area in frogs. The ability to measure fine-scale changes or alterations in parasite distributions has the potential to provide further insight about ecological questions concerning habitat preference, resource selection, host pathology and disease control.

Measures of spatial heterogeneity in the liver tissue micro-environment as predictive factors for fibrosis score

The organization and interaction between hepatocytes and other hepatic non-parenchymal cells plays a pivotal role in maintaining normal liver function and structure. Although spatial heterogeneity within the tumor micro-environment has been proven to be a fundamental feature in cancer progression, the role of liver tissue topology and micro-environmental factors in the context of liver damage in chronic infection has not been widely studied yet. We obtained images from 110 core needle biopsies from a cohort of chronic hepatitis B patients with different fibrosis stages according to METAVIR score. The tissue sections were immunofluorescently stained and imaged to determine the locations of CD45 positive immune cells and HBsAg-negative and HBsAg-positive hepatocytes within the tissue. We applied several descriptive techniques adopted from ecology, including Getis-Ord, the Shannon Index and the Morisita-Horn Index, to quantify the extent to which immune cells and different types of liver cells co-localize in the tissue biopsies. Additionally, we modeled the spatial distribution of the different cell types using a joint log-Gaussian Cox process and proposed several features to quantify spatial heterogeneity. We then related these measures to the patient fibrosis stage by using a linear discriminant analysis approach. Our analysis revealed that the co-localization of HBsAg-negative hepatocytes with immune cells and the co-localization of HBsAg-positive hepatocytes with immune cells are equally important factors for explaining the METAVIR score in chronic hepatitis B patients. Moreover, we found that if we allow for an error of 1 on the METAVIR score, we are able to reach an accuracy of around 80%. With this study we demonstrate how methods adopted from ecology and applied to the liver tissue micro-environment can be used to quantify heterogeneity and how these approaches can be valuable in biomarker analyses for liver topology.

High-yield areas to grade tumor budding in colorectal cancer: A practical approach for pathologists

BACKGROUND

Tumor budding (TB) has significant prognostic implication in stage II colorectal cancer (CRC) and is graded based on the International Tumor Budding Consensus Conference (ITBCC) protocol. In the current study, we evaluate tumor budding and its relationship to multiple histologic features in 104 tumors.

METHODS

One-hundred four resected CRC cases were retrieved. Tumor bud count and TB grade were compared to the final tumor bud count/TB grade of the tumor per ITBCC protocol. The following high-yield co-features were assessed in each slide: highest T stage, presence of benign mucosa, presence of a precursor lesion, and highest tumor volume.

RESULTS

Twenty-nine (28 %) cases had discrepancies between slide TB grade and final TB grade. The least discrepancies were seen in slides with benign mucosa (7 %) and precursor lesions (7 %). Among stage II patients without high-risk features, no discrepancies were observed in slides with benign mucosa. Slides with deepest invasion (r_s = 1.000, p = 0.01) and benign mucosa (r_s = 0.957, p < 0.001) had the strongest correlation with final tumor bud count in the same stage II subgroup. Similar relationships were observed when comparing final TB grade. Deepest invasion, tumor volume, as well as lymphovascular invasion, when present, also showed strong correlations with final TB grade in the entire cohort (r_s = 0.828-0.845, p < 0.001).

CONCLUSION

Our study is the first study to evaluate the relationship between TB grade and co-existing histologic features. We highlight the benefit of focusing on slides with high-yield co-features, with the strongest correlation seen in slides with adjacent benign mucosa and precursor lesions.

Data synthesis and adversarial networks: A review and meta-analysis in cancer imaging

Despite technological and medical advances, the detection, interpretation, and treatment of cancer based on imaging data continue to pose significant challenges. These include inter-observer variability, class imbalance, dataset shifts, inter- and intra-tumour heterogeneity, malignancy determination, and treatment effect uncertainty. Given the recent advancements in image synthesis, Generative Adversarial Networks (GANs), and adversarial training, we assess the potential of these technologies to address a number of key challenges of cancer imaging. We categorise these challenges into (a) data scarcity and imbalance, (b) data access and privacy, (c) data annotation and segmentation, (d) cancer detection and diagnosis, and (e) tumour profiling, treatment planning and monitoring. Based on our analysis of 164 publications that apply adversarial training techniques in the context of cancer imaging, we highlight multiple underexplored solutions with research potential. We further contribute the Synthesis Study Trustworthiness Test (SynTRUST), a meta-analysis framework for assessing the validation rigour of medical image synthesis studies. SynTRUST is based on 26 concrete measures of thoroughness, reproducibility, usefulness, scalability, and tenability. Based on SynTRUST, we analyse 16 of the most promising cancer imaging challenge solutions and observe a high validation rigour in general, but also several desirable improvements. With this work, we strive to bridge the gap between the needs of the clinical cancer imaging community and the current and prospective research on data synthesis and adversarial networks in the artificial intelligence community.

Caveat Medicus: It's Time to Re-Think Stratification, You May Not Be Helping

Background

The focus of the present Letter is on the large and seemingly fertile body of work captured under the umbrella of 'patient stratification'.

Objectives

I identify and explain a fundamental methodological flaw underlying the manner in which the development of an increasingly large number of new stratification strategies is approached.

Design

I show an inherent conflict between the assumptions made, and the very purpose of stratification and its application in practice.

Methods

I analyse the methodological underpinnings of stratification as presently done and draw parallels with conceptually similarly flawed precedents which are now widely recognized.

Results

The highlighted flaw is shown to undermine the overarching ultimate goal of improved patient outcomes by undue fixation on an ill-founded proxy.

Conclusion

I issue a call for a re-think of the problem and the processes leading to the adoption of new stratification strategies in the clinic.

Deep-learning-based classification of desmoplastic reaction on H&E predicts poor prognosis in oesophageal squamous cell carcinoma

AIMS

Desmoplastic reaction (DR) categorisation has been shown to be a promising prognostic factor in oesophageal squamous cell carcinoma (ESCC). The usual DR evaluation is performed using semiquantitative scores, which can be subjective. This study aimed to investigate whether a deep-learning classifier could be used for DR classification. We further assessed the prognostic significance of the deep-learning classifier and compared it to that of manual DR reporting and other pathological factors currently used in the clinic.

METHODS AND RESULTS

From 222 surgically resected ESCC cases, 31 randomly selected haematoxylin-eosin-digitised whole slides of patients with immature DR were used to train and develop a deep-learning classifier. The classifier was trained for 89 370 iterations. The accuracy of the deep-learning classifier was assessed to 30 unseen cases, and the results revealed a Dice coefficient score of 0.81. For survival analysis, the classifier was then applied to the entire cohort of patients, which was split into a training (n = 156) and a test (n = 66) cohort. The automated DR classification had a higher prognostic significance for disease-specific survival than the manually classified DR in both the training and test cohorts. In addition, the automated DR classification outperformed the prognostic accuracy of the gold-standard factors of tumour depth and lymph node metastasis.

CONCLUSIONS

This study demonstrated that DR can be objectively and quantitatively assessed in ESCC using a deep-learning classifier and that automatically classed DR has a higher prognostic significance than manual DR and other features currently used in the clinic.

Proposal of an automated tumor-stromal ratio assessment algorithm and a nomogram for prognosis in early-stage invasive breast cancer

BACKGROUND

The tumor-stromal ratio (TSR) has been verified to be a prognostic factor in many solid tumors. In most studies, it was manually assessed on routinely stained H&E slides. This study aimed to assess the TSR using image analysis algorithms developed by the Qupath software, and integrate the TSR into a nomogram for prediction of the survival in invasive breast cancer (BC) patients.

METHODS

A modified TSR assessment algorithm based on the recognition of tumor and stroma tissues was developed using the Qupath software. The TSR of 234 invasive BC specimens in H&E-stained tissue microarrays (TMAs) were assessed with the algorithm and categorized as stroma-low or stroma-high. The consistency of TSR estimation between Qupath prediction and pathologist annotation was analyzed. Univariable and multivariable analyses were applied to select potential risk factors and a nomogram for predicting survival in invasive BC patients was constructed and validated. An extra TMA containing 110 specimens was obtained to validate the conclusion as an independent cohort.

RESULTS

In the discovery cohort, stroma-low and stroma-high were identified in 43.6% and 56.4% cases, respectively. Good concordance was observed between the pathologist annotated and Qupath predicted TSR. The Kaplan-Meier curve showed that stroma-high patients were associated with worse 5-DFS compared to stroma-low patients (p = 0.007). Multivariable analysis identified age, T stage, N status, histological grade, ER status, HER-2 gene, and TSR as potential risk predictors, which were included in the nomogram. The nomogram was well calibrated and showed a favorable predictive value for the recurrence of BC. Kaplan-Meier curves showed that the nomogram had a better risk stratification capability than the TNM staging system. In the external validation of the nomogram, the results were further validated.

CONCLUSIONS

Based on H&E-stained TMAs, this study successfully developed image analysis algorithms for TSR assessment and constructed a nomogram for predicting survival in invasive BC.

COVID-19 and Science Communication: The Recording and Reporting of Disease Mortality

The ongoing COVID-19 pandemic has brought science to the fore of public discourse and, considering the complexity of the issues involved, with it also the challenge of effective and informative science communication. This is a particularly contentious topic, in that it is both highly emotional in and of itself; sits at the nexus of the decision-making process regarding the handling of the pandemic, which has effected lockdowns, social behaviour measures, business closures, and others; and concerns the recording and reporting of disease mortality. To clarify a point that has caused much controversy and anger in the public debate, the first part of the present article discusses the very fundamentals underlying the issue of causative attribution with regards to mortality, lays out the foundations of the statistical means of mortality estimation, and concretizes these by analysing the recording and reporting practices adopted in England and their widespread misrepresentations. The second part of the article is empirical in nature. I present data and an analysis of how COVID-19 mortality has been reported in the mainstream media in the UK and the USA, including a comparative analysis both across the two countries as well as across different media outlets. The findings clearly demonstrate a uniform and worrying lack of understanding of the relevant technical subject matter by the media in both countries. Of particular interest is the finding that with a remarkable regularity (ρ>0.998), the greater the number of articles a media outlet has published on COVID-19 mortality, the greater the proportion of its articles misrepresented the disease mortality figures.

Spatial heterogeneity of immune infiltration predicts the prognosis of nasopharyngeal carcinoma patients

Spatial information on the tumor immune microenvironment is of clinical relevance. Here, we aimed to quantify the spatial heterogeneity of lymphocytes and cancer cells and evaluated its prognostic value in patients with nasopharyngeal carcinoma (NPC). The scanned immunohistochemistry images of 336 NPC patients from two different hospitals were used to generate cell density maps for tumor and immune cells. Then, Getis-Ord hotspot analysis, a spatial statistic method used to describe species biodiversity in ecological habitats, was applied to identify cancer, immune, and immune-cancer hotspots. The results showed that cancer hotspots were not associated with any of the studied clinical outcomes, while immune-cancer hotspots predicted worse overall survival (OS) in the training cohort. In contrast, a high immune hotspot score was significantly associated with better OS (HR 0.41, 95% CI 0.22–0.77, P = .006), disease-free survival (DFS) (HR 0.43, 95% CI 0.24–0.75, P = .003) and distant metastasis-free survival (DMFS) (HR 0.40, 95% CI 0.20–0.81, P = .011) in NPC patients in the training cohort, and similar associations were also evident in the validation cohort. Importantly, multivariate analysis revealed that the immune hotspot score remained an independent prognostic indicator for OS, DFS, and DMFS in both cohorts. We explored the spatial heterogeneity of cancer cells and lymphocytes in the tumor microenvironment of NPC patients using digital pathology and ecological analysis methods and further constructed three spatial scores. Our study demonstrates that spatial variation may aid in the identification of the clinical prognosis of NPC patients, but further investigation is needed.

Two Ensemble-CNN Approaches for Colorectal Cancer Tissue Type Classification

In recent years, automatic tissue phenotyping has attracted increasing interest in the Digital Pathology (DP) field. For Colorectal Cancer (CRC), tissue phenotyping can diagnose the cancer and differentiate between different cancer grades. The development of Whole Slide Images (WSIs) has provided the required data for creating automatic tissue phenotyping systems. In this paper, we study different hand-crafted feature-based and deep learning methods using two popular multi-classes CRC-tissue-type databases: Kather-CRC-2016 and CRC-TP. For the hand-crafted features, we use two texture descriptors (LPQ and BSIF) and their combination. In addition, two classifiers are used (SVM and NN) to classify the texture features into distinct CRC tissue types. For the deep learning methods, we evaluate four Convolutional Neural Network (CNN) architectures (ResNet-101, ResNeXt-50, Inception-v3, and DenseNet-161). Moreover, we propose two Ensemble CNN approaches: Mean-Ensemble-CNN and NN-Ensemble-CNN. The experimental results show that the proposed approaches outperformed the hand-crafted feature-based methods, CNN architectures and the state-of-the-art methods in both databases.